/*Replication do file for: Songying Fang and Fanglu Sun. Gauging Chinese Public Support for China’s Role in Peacekeeping. CJIP (2019)*/


/******************************************************************************/
/******************************************************************************/
/******************************************************************************/
/*Data source for Figure 1 and 2: The International Peace Institute Peacekeeping Database */
/*Fig. 1. UN Peacekeeping Troop Contributions by the P5, 1990-2015*/
clear
insheet using "Country-level-data.csv",clear  /*it totals the contribution of the country over all missions*/
br
br date troopcontributions policecontributions totalcontributions if contributoriso3=="CHN"

ds date-totalcontributionssd, has(type string)
foreach var of varlist `r(varlist)' {
        replace `var' = "" if strpos(`var',"NA")
        destring `var', replace
    }
    
gen mydate=date(date, "YMD")
format mydate %tdCY-N-D


/*P5*/
preserve 
    *keep if contributoriso3 =="USA"
    encode contributoriso3, gen(contributoriso3_numeric)
    *tsset contributoriso3_numeric mydate, monthly
    tsset contributoriso3_numeric mydate   
    tsline totalcontributions if contributoriso3 =="CHN" || ///
    tsline totalcontributions if contributoriso3 =="RUS" || ///
    tsline totalcontributions if contributoriso3 =="FRA" || ///
    tsline totalcontributions if contributoriso3 =="GBR" || ///
    tsline totalcontributions if contributoriso3 =="USA", /// 
             ytitle("Monthly Personnel Contribution", height(5)) ///
             xtitle("") xla(, labsize(small)) yla(, labsize(small)) ///
             legend( pos(5) row(1) lab(1 "China")  lab(2 "Russia") lab(3 "France") lab(4 "UK") lab(5 "USA" ))
    graph export "personnelcontribution_p5.pdf",replace
restore

/******************************************************************************/
/*Fig. 2. UN Peacekeeping Financial Contributions by the P5 (Nominal USD), 1994-2015*/
insheet using "Financial-data.csv",clear
rename v25 assessedadjusted2010monthly
   
gen mydate=date(date, "MDY")
*format mydate %dM_d,_CY
format mydate %td

gen assessedcontributionactualyearly=assessedcontributionactual/1000000000 
sum assessedcontributionactualyearly

preserve 
    keep year tcciso3alpha assessedcontributionactualyearly
	keep  if tcciso3alpha =="CHN" | tcciso3alpha =="RUS" |  tcciso3alpha =="FRA" |tcciso3alpha =="GBR" | tcciso3alpha =="USA"   
    duplicates report
	duplicates drop 
	line assessedcontributionactualyearly year if tcciso3alpha =="CHN" || ///
	line assessedcontributionactualyearly year if tcciso3alpha =="RUS" || ///
	line assessedcontributionactualyearly year if tcciso3alpha =="FRA" || ///
	line assessedcontributionactualyearly year if tcciso3alpha =="GBR" || ///
	line assessedcontributionactualyearly year if tcciso3alpha =="USA" || , ///
	ytitle("Yearly Financial Contribution (Billion Dollars)", height(5)) ///
	xtitle("") xla(, labsize(small)) ///
	legend( pos(5) row(1) lab(1 "China")  lab(2 "Russia") lab(3 "France") lab(4 "UK") lab(5 "USA" ))
	graph export "financialcontribution_p5.pdf",replace
restore


/******************************************************************************/
/******************************************************************************/
/******************************************************************************/
/*Data Source for  Figure 3 onward : Survey Conducted by Authors*/
/*open data*/
use China_Peacekeeping_Chinese.dta,clear

/******************************************************************************/
/*Figure 3: China’s Interests in the Host Country and Average Level of Chinese Public Support for PKO*/
/*kwallis Test*/
kwallis support , by(interests) /* p.=.673*/
kwallis supportdummy , by(interests) /*p=.827*/

ttest supportdummy, by (humanitarian)
matrix mat01 = (r(mu_2), r(mu_2)-1.96*(r(sd_2)/sqrt(r(N_2))), r(mu_2)+1.96*(r(sd_2)/sqrt(r(N_2))),r(p), r(mu_1))

ttest supportdummy if interests==1 | interests==2 , by (economic)   // humanistarian vs. economic
matrix mat02 = (r(mu_2), r(mu_2)-1.96*(r(sd_2)/sqrt(r(N_2))), r(mu_2)+1.96*(r(sd_2)/sqrt(r(N_2))),r(p), r(mu_1))

ttest supportdummy if interests==1 | interests==3 , by (strategic)  // humanistarian vs. strategic
matrix mat03 = (r(mu_2), r(mu_2)-1.96*(r(sd_2)/sqrt(r(N_2))), r(mu_2)+1.96*(r(sd_2)/sqrt(r(N_2))),r(p), r(mu_1))

ttest supportdummy  if interests==1 | interests==4 , by (terrorism) // humanistarian vs. security
matrix mat04 = (r(mu_2), r(mu_2)-1.96*(r(sd_2)/sqrt(r(N_2))), r(mu_2)+1.96*(r(sd_2)/sqrt(r(N_2))),r(p),r(mu_1))

capture drop t1 t2 t3 t4 diff
matrix t1 = mat01\mat02\mat03\mat04 /*humanitarian vs. economic vs. strategic vs. terrorism*/
matrix rownames t1 = a b c d
matrix list t1
display t1[1,4] /*display p-value*/
display t1[2,4]
display t1[3,4]
display t1[4,4]

local pp1=abs(round(t1[1,4], .001))
local pp2=abs(round(t1[2,4], .001))
local pp3=abs(round(t1[3,4], .001))
local pp4=abs(round(t1[4,4], .001))

local diff1=abs(round(t1[1,1]-t1[1,5], .001))
local diff2=abs(round(t1[2,1]-t1[2,5], .001))
local diff3=abs(round(t1[3,1]-t1[3,5], .001))
local diff4=abs(round(t1[4,1]-t1[4,5], .001))

coefplot (matrix(t1[.,1]), ci("t1[.,2] t1[.,3]")  msymbol(D) color(black) ciopts(lwidth(*2) lpatt(solid) lcol(blue))), ///
coeflabel(a = `" "Humanitarian Crisis" "' ///
 		   b = `" "Economic Interest" "' ///
           c = `" "Strategic Interest" "'  ///
		   d = `" "Security Interest" "' )  ///
mlabel format(%9.3g) mlabposition(1) ///
xtitle(Proportion of Support) cismooth 
graph export fttestinterestssimple.eps, as(eps) preview(on) replace	   
graph save fttestinterestssimple,replace 

/******************************************************************************/
/*Fig. 4. IO Authorisation and Average Level of Chinese Public Support for PKO. */
/*kwallis Test*/
kwallis support , by(IO) /* p=.9621*/
kwallis supportdummy , by(IO) /*p=.5529*/

ttest supportdummy, by (UN) 
return list
matrix mat05 = (r(mu_2), r(mu_2)-1.96*(r(sd_2)/sqrt(r(N_2))), r(mu_2)+1.96*(r(sd_2)/sqrt(r(N_2))),r(p),r(mu_1))

ttest supportdummy, by (AU) 
matrix mat06 = (r(mu_2), r(mu_2)-1.96*(r(sd_2)/sqrt(r(N_2))), r(mu_2)+1.96*(r(sd_2)/sqrt(r(N_2))),r(p),r(mu_1))

capture drop t1 t2  diff*
matrix t1 = mat05\mat06 /*humanitarian vs. economic vs. strategic vs. terrorism*/
matrix rownames t1 = a b 
matrix list t1
display t1[1,4] /*display p-value*/
display t1[2,4]

local pp1=abs(round(t1[1,4], .001))
local pp2=abs(round(t1[2,4], .001))

local diff1=round(t1[1,1]-t1[1,5], .001)
display `diff1'
local diff2=abs(round(t1[2,1]-t1[2,5], .001))
display `diff2'

coefplot (matrix(t1[.,1]), ci("t1[.,2] t1[.,3]") msymbol(D) color(black) ciopts(lwidth(*2) lpatt(solid) lcol(blue))), ///
coeflabel(a = `" "UN" "' ///
 		   b = `" "AU" "') ///
mlabel format(%9.3g) mlabposition(1) ///
xtitle(Proportion of Support) cismooth 
graph export fttestIOsimple.eps, as(eps) preview(on) replace	   
graph save fttestIOsimple,replace 

/******************************************************************************/
/*Fig. 5. Chinese Public Support for Three Types of PKO.  */
kwallis support , by(PKOtype) /*for Figure fPKOinterestsordinal.eps, p.=.0001*/
kwallis supportdummy , by(PKOtype) /*for Figure fPKOinterests.eps, p=.0001*/

codebook PKOtype
ttest supportdummy, by (financial) /*financial vs. non-financial*/
return list
matrix mat07 = (r(mu_2), r(mu_2)-1.96*(r(sd_2)/sqrt(r(N_2))), r(mu_2)+1.96*(r(sd_2)/sqrt(r(N_2))),r(p),r(mu_1))

ttest supportdummy if PKOtype==1 | PKOtype==2 , by (personnel) /*financial vs. personnel*/
matrix mat08 = (r(mu_2), r(mu_2)-1.96*(r(sd_2)/sqrt(r(N_2))), r(mu_2)+1.96*(r(sd_2)/sqrt(r(N_2))),r(p),r(mu_1))

ttest supportdummy if PKOtype==1 | PKOtype==3, by (militaryleadership) /*financial vs. militaryleadership*/
matrix mat09 = (r(mu_2), r(mu_2)-1.96*(r(sd_2)/sqrt(r(N_2))), r(mu_2)+1.96*(r(sd_2)/sqrt(r(N_2))),r(p),r(mu_1))

capture drop t1 t2 t3 t4 diff
matrix t1 = mat07\mat08\mat09 /*humanitarian vs. economic vs. strategic vs. terrorism*/
matrix rownames t1 = a b c 
matrix list t1
display t1[1,4] /*display p-value*/
display t1[2,4]
display t1[3,4]

local pp1=abs(round(t1[1,4], .001)) /*extremelysmall*/
local pp2=abs(round(t1[2,4], .001))
local pp3=abs(round(t1[3,4], .001))

local diff1=round(t1[1,1]-t1[1,5], .001)
local diff2=round(t1[2,1]-t1[2,5], .001)
local diff3=round(t1[3,1]-t1[3,5], .001)

coefplot (matrix(t1[.,1]), ci("t1[.,2] t1[.,3]")  msymbol(D) color(black) ciopts(lwidth(*2) lpatt(solid) lcol(blue))), ///
coeflabel(a = `" "Financial Contribution" "' ///
 		   b = `" "Personnel Contribution" "' ///
		   c = `" "Military Leadership" "' )  ///
mlabel format(%9.3g) mlabposition(1) ///
xtitle(Proportion of Support) cismooth 
graph export fttestPKOtypesimple.eps, as(eps) preview(on) replace	   
graph save fttestPKOtypesimple,replace 

/******************************************************************************/
/*Fig. 6. Chinese Public Support for Different Types of PKO in Different Scenarios.*/
/*support for financial contribution*/
ttest supportdummy if financial==1, by (UN) 
return list
matrix mat1 = (r(mu_2), r(mu_2)-1.96*(r(sd_2)/sqrt(r(N_2))), r(mu_2)+1.96*(r(sd_2)/sqrt(r(N_2))),r(p))

ttest supportdummy if financial==1, by (AU) 
matrix mat2 = (r(mu_2), r(mu_2)-1.96*(r(sd_2)/sqrt(r(N_2))), r(mu_2)+1.96*(r(sd_2)/sqrt(r(N_2))),r(p))

ttest supportdummy if financial==1, by (humanitarian) 
matrix mat3 = (r(mu_2), r(mu_2)-1.96*(r(sd_2)/sqrt(r(N_2))), r(mu_2)+1.96*(r(sd_2)/sqrt(r(N_2))),r(p))

ttest supportdummy if financial==1, by (economic) 
matrix mat4 = (r(mu_2), r(mu_2)-1.96*(r(sd_2)/sqrt(r(N_2))), r(mu_2)+1.96*(r(sd_2)/sqrt(r(N_2))),r(p))

ttest supportdummy if financial==1, by (strategic) 
matrix mat5 = (r(mu_2), r(mu_2)-1.96*(r(sd_2)/sqrt(r(N_2))), r(mu_2)+1.96*(r(sd_2)/sqrt(r(N_2))),r(p))

ttest supportdummy if financial==1, by (terrorism) 
matrix mat6 = (r(mu_2), r(mu_2)-1.96*(r(sd_2)/sqrt(r(N_2))), r(mu_2)+1.96*(r(sd_2)/sqrt(r(N_2))),r(p))


/*support for personnel contribution*/
ttest supportdummy if personnel ==1, by (UN) 
matrix mat7 = (r(mu_2), r(mu_2)-1.96*(r(sd_2)/sqrt(r(N_2))), r(mu_2)+1.96*(r(sd_2)/sqrt(r(N_2))),r(p))

ttest supportdummy if personnel ==1, by (AU) 
matrix mat8 = (r(mu_2), r(mu_2)-1.96*(r(sd_2)/sqrt(r(N_2))), r(mu_2)+1.96*(r(sd_2)/sqrt(r(N_2))),r(p))

ttest supportdummy if personnel ==1, by (humanitarian) 
matrix mat9 = (r(mu_2), r(mu_2)-1.96*(r(sd_2)/sqrt(r(N_2))), r(mu_2)+1.96*(r(sd_2)/sqrt(r(N_2))),r(p))

ttest supportdummy if personnel ==1, by (economic) 
matrix mat10 = (r(mu_2), r(mu_2)-1.96*(r(sd_2)/sqrt(r(N_2))), r(mu_2)+1.96*(r(sd_2)/sqrt(r(N_2))),r(p))

ttest supportdummy if personnel ==1, by (strategic) 
matrix mat11 = (r(mu_2), r(mu_2)-1.96*(r(sd_2)/sqrt(r(N_2))), r(mu_2)+1.96*(r(sd_2)/sqrt(r(N_2))),r(p))

ttest supportdummy if personnel ==1, by (terrorism) 
matrix mat12 = (r(mu_2), r(mu_2)-1.96*(r(sd_2)/sqrt(r(N_2))), r(mu_2)+1.96*(r(sd_2)/sqrt(r(N_2))),r(p))

/*support for military leadership*/
ttest supportdummy if militaryleadership ==1, by (UN) 
matrix mat13 = (r(mu_2), r(mu_2)-1.96*(r(sd_2)/sqrt(r(N_2))), r(mu_2)+1.96*(r(sd_2)/sqrt(r(N_2))),r(p))

ttest supportdummy if militaryleadership ==1, by (AU) 
matrix mat14 = (r(mu_2), r(mu_2)-1.96*(r(sd_2)/sqrt(r(N_2))), r(mu_2)+1.96*(r(sd_2)/sqrt(r(N_2))),r(p))

ttest supportdummy if militaryleadership ==1, by (humanitarian) 
matrix mat15 = (r(mu_2), r(mu_2)-1.96*(r(sd_2)/sqrt(r(N_2))), r(mu_2)+1.96*(r(sd_2)/sqrt(r(N_2))),r(p))

ttest supportdummy if militaryleadership ==1, by (economic) 
matrix mat16 = (r(mu_2), r(mu_2)-1.96*(r(sd_2)/sqrt(r(N_2))), r(mu_2)+1.96*(r(sd_2)/sqrt(r(N_2))),r(p))

ttest supportdummy if militaryleadership ==1, by (strategic) 
matrix mat17 = (r(mu_2), r(mu_2)-1.96*(r(sd_2)/sqrt(r(N_2))), r(mu_2)+1.96*(r(sd_2)/sqrt(r(N_2))),r(p))

ttest supportdummy if militaryleadership ==1, by (terrorism) 
matrix mat18 = (r(mu_2), r(mu_2)-1.96*(r(sd_2)/sqrt(r(N_2))), r(mu_2)+1.96*(r(sd_2)/sqrt(r(N_2))),r(p))

capture drop t1 t2 t3
matrix t1 = mat1\mat2\mat3\mat4\mat5\mat6
matrix rownames t1 = a b c d e f

matrix t2 = mat7\mat8\mat9\mat10\mat11\mat12
matrix rownames t2 = a b c d e f

matrix t3 = mat13\mat14\mat15\mat16\mat17\mat18
matrix rownames t3 = a b c d e f

coefplot (matrix(t1[.,1]), ci("t1[.,2] t1[.,3]") label(Financial Contribution) color(black) ciopts(lwidth(*2) lpatt(dot) lcol(black)))  ///
 (matrix(t2[.,1]), ci("t2[.,2] t2[.,3]") label(Personnel Contribution) lcol(blue )  msymbol(S) color(blue) ciopts(lwidth(*2)lpatt(dash)lcol(blue)))  ///
 (matrix(t3[.,1]), ci("t3[.,2] t3[.,3]") label(Military Leadership) lcol(red)  msymbol(D) color(red) ciopts(lwidth(*2) lpatt(solid)lcol(red))) ,  ///
 coeflabel(a = `" "UN" "' ///
 		   b = `" "AU" "' ///
           c = `" "Humanitarian Crisis" "' ///
           d = `" "Economic Interest" "' ///
           e = `" "Stategic Interest" "' ///
           f = `" "Security Interest" "') ///
mlabel format(%9.3g) mlabposition(1) ///
xtitle(Proportion of Support) cismooth  

/******************************************************************************/
/* Fig. 7. Logistic Analysis of Support for China’s Participation in Peacekeeping.*/
gen securityXpersonnel=terrorism*personnel
gen securityXmilitaryleadership=terrorism*militaryleadership
gen UNXpersonnel=UN*personnel
gen UNXmilitaryleadership=UN*militaryleadership
label variable securityXpersonnel  "Security × Personnel"
label variable securityXmilitaryleadership "Security × Military Leadership"
label variable UNXpersonnel "UN × Personnel"
label variable UNXmilitaryleadership "UN × Military Leadership"

global treatments "economic strategic terrorism UN personnel militaryleadership"
global policovariates "improveimage improvebirelations responsivepower hostpermission nationalism  Chinastatus Chinaecoprospect  news "
global sociacovaraites "age han male eastern central rural college SOE ccp income socialstatus"
global interactions "securityXpersonnel securityXmilitaryleadership UNXpersonnel UNXmilitaryleadership"

logit supportdummy  $treatments $interactions $policovariates  $sociacovaraites
estimates store poolinteraction

coefplot poolinteraction, xline(0) drop(_cons) byopts(row(1))  xtitle("Coefficient with 95% Confidence Interval")  ///
          mcolor(black) ciopts(lwidth(*2 ..) lcolor(blue . .))  ///
          headings(economic = "{bf:Contextual Variables}" securityXpersonnel="{bf:Interaction Terms}" improveimage ="{bf:Attitudes and Perceptions}" age ="{bf:Socio-Demographics} ", labcolor(orange))  ///
		  groups(economic strategic terrorism = "{bf:}"  ///
		  personnel militaryleadership  = "{bf:}" ///	 
		  improveimage improvebirelations responsivepower = "{bf:}",labsize(3)labgap(50) )  ///
		  coeflabels(,notick labsize(2.5) labcolor(black) labgap(1)) graphregion(margin(l=0)) 
graph export fmodelpoolinteraction.eps, as(eps) preview(on) replace	   
graph save fmodelpoolinteraction,replace  

/******************************************************************************/
/*Fig. 8. Generational Differences in the Predicted Level of Public Support for China’s Participation in PKOs*/
margins, at(age=(1 (1) 6)) atmeans vsquish
marginsplot, xlabel( 1 "<=20" 2 "20-30" 3 "30-40" 4 "40-50" 5 "50-60" 6 ">60", angle(45))  xlabel(,labsize(medium)) ///
      ytitle("Predicted Level of Public Support")  title("")
graph export fpredage.eps, as(eps) preview(on) replace	   
graph save fpredage,replace

/******************************************************************************/
/**************************Online Appendix**************************************/
/******************************************************************************/
**************************************
*  Section A: Descriptive Statistics and Variable Measurements
**************************************
/*Table1. Descriptive statistics of the respondents*/
local vars age han male eastern central rural college SOE ccp income socialstatus  improveimage improvebirelations responsivepower hostpermission nationalism  Chinastatus Chinaecoprospect news
estpost su `vars', detail
est store desc
esttab desc using tdesc.tex, replace ///
   nomtitles ///
   collabels(\multicolumn{1}{c}{{Observation}} \multicolumn{1}{c}{{Mean}} \multicolumn{1}{c}{{Median}} \multicolumn{1}{l}{{Min}} \multicolumn{1}{l}{{Max}}) ///
   cells("count mean p50 min max") label booktabs nonum  gaps noobs  ///
   title (\label{tdesc} Descriptive statistics of the sociodemographic data of the respondents) 

/*Table 2. Distribution of the Treatments.*/
tabulate interests IO if PKOtype==1 
tabulate interests IO if PKOtype==2
tabulate interests IO if PKOtype==2

*********************************************
*  Section C: Randomization Checks
**********************************************
/*Table 3. Randomization Checks */
/*Two Treatments of Authorization*/
tab IO
estpost su  $covariates if IO==0
est store IO0
estpost su  $covariates if IO==1
est store IO1

tab IO age, chi2
tab IO han, chi2
tab IO male, chi2
tab IO eastern, chi2
tab IO central, chi2
tab IO rural, chi2
tab IO college, chi2
tab IO SOE, chi2
tab IO ccp, chi2
tab IO income, chi2
tab IO socialstatus, chi2
tab IO news, chi2
tab IO nationalism, chi2
  
/*Four Treatments of Interests */
tab interests
estpost su  $covariates if interests ==1
est store interests1
estpost su  $covariates if interests ==2
est store interests2
estpost su  $covariates if interests ==3
est store interests3
estpost su  $covariates if interests ==4
est store interests4

tab interests age, chi2 
tab interests han, chi2 
tab interests male, chi2
tab interests eastern, chi2
tab interests central, chi2
tab interests rural, chi2
tab interests college, chi2
tab interests SOE, chi2
tab interests ccp, chi2
tab interests income, chi2
kwallis socialstatus, by(interests) 
tab interests socialstatus, chi2 
kwallis news, by(interests)
tab interests news, chi2 
tab interests nationalism, chi2

/*Three Treatments of Participation Type*/
tab PKOtype
estpost su  $covariates if PKOtype ==1
est store PKOtype1
estpost su  $covariates if PKOtype ==2
est store PKOtype2
estpost su  $covariates if PKOtype ==3
est store PKOtype3

tab PKOtype age, chi2 
tab PKOtype han, chi2
tab PKOtype male, chi2
tab PKOtype eastern, chi2
tab PKOtype central, chi2
tab PKOtype rural, chi2
tab PKOtype college, chi2
tab PKOtype SOE, chi2
tab PKOtype ccp, chi2
tab PKOtype income, chi2
tab PKOtype socialstatus, chi2
tab PKOtype news, chi2
tab PKOtype nationalism, chi2

/*Table 3 in the Appendix*/
esttab IO0 IO1 interests1 interests2 interests3 interests4 PKOtype1 PKOtype2 PKOtype3 using tcovariatebalanced.tex, ///
  mtitle("UN" "AU""Humanitarian" "Economic" "Strategic"  "Security" "Financial" "Personnel" "Military Leadership" ) ///
  replace cells( mean(fmt(2)) ) label booktabs nonum collabels(none) gaps noobs ///
  title (\label{tcovariatebalanced.tex} Randomization Check/Balance Test) 

*********************************************
*  Section D: Randomization Checks
**********************************************
logit supportdummy $treatments  $policovariates  $sociacovaraites if PKOtype ==1
estimates store Financial

logit supportdummy $treatments  $policovariates  $sociacovaraites if PKOtype ==2
estimates store Personnel

logit supportdummy $treatments  $policovariates   $sociacovaraites if PKOtype ==3
estimates store MilitaryLeadership

/*Fig.1. Logit Models of China’s Public Support for Peacekeeping.*/
coefplot Financial ||  Personnel ||  MilitaryLeadership, xline(0) drop(_cons) byopts(row(1))  xtitle("Coefficient with 95% Confidence Interval")  ///
          mcolor(black) ciopts(lwidth(*2 ..) lcolor(blue . .))  ///
          headings(economic = "{bf:Contextual Variables}" improveimage ="{bf:Attitudes and Perceptions}" age ="{bf:Socio-Demographics}", labcolor(orange))  ///
		  groups(economic strategic terrorism = "{bf:}"  ///
		  personnel militaryleadership  = "{bf:}" ///	 
		  improveimage improvebirelations responsivepower = "{bf:}",labsize(3)labgap(50) )  ///
		  coeflabels(,notick labsize(2.5) labcolor(black) labgap(1)) graphregion(margin(l=0)) 
graph export fmodelsep.eps, as(eps) preview(on) replace	   
graph save fmodelsep,replace 

/*Table 4. Logit Models of China’s Public Support for Peacekeeping*/
esttab Financial Personnel MilitaryLeadership using tmodelsep.tex, replace ///
         b(%10.3f) se scalars("ll Log lik." "chi2 Chi-squared") ///
         label mtitles 		  
